Avalanche Photodiodes Performance Parameters Estimation...
Transcript of Avalanche Photodiodes Performance Parameters Estimation...
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1822
www.ijsetr.org
Avalanche Photodiodes Performance Parameters
Estimation under Thermal Irradiation Fields
Ahmed Nabih Zaki Rashed1, Abd El-Naser A. Mohamed
1, Imbaby I. Mahmoud
2,
Mohamed S. El_Tokhy2, and Osama H. Elgzar
2
1Electronics and Electrical Communication Engineering Department,
Faculty of Electronic Engineering, Menouf 32951, Menoufia University, EGYPT 2Engineering Department, Nuclear Research Center, Atomic Energy Authority, P.O. 13759, Inshas, EGYPT
Abstract— Radiation-induced damage in Avalanche
Photodiode (APD) was shown to result from the dark current and
a change of the effective doping concentration occurring within
the photodiodes. In this paper a model to reveals the effect of
ionizing radiation and temperature on the performance of APDs
is built by using Vissim environment. This proposed model
provides a mean to control the properties of APD when they are
selected to operate in thermal radiation environments. Efficiency,
sensitivity, responsitivity, detectivity, noise equivalent power,
excess noise factor, normalized detectivity, and signal to noise
ratio are modeled. The temperature effects are combined with
radiation effects to formulate a rigours treatment for the APD
behavior. The results are validated against published
experimental work in temperature case and show good
agreement.
Index Terms— APD, Radiation effects, Excess noise factor,
Noise equivalent power, Responsitivity, and Detectvity .
I. INTRODUCTION
Optoelectronic components have become important elements in modern electronic systems due to the many intrinsic advantages
of optical signal transfer, especially the large available bandwidths, high immunity to electromagnetic interference,
and light weight [1]-[6]. Among of these components is the photodiode which when is exposed to the effects of light, then
the photon energy is used to break the covalent bond releasing electrons and creating a hole in the process. If the generation of
electrons and holes occurred in the depletion area, then the existing electric field removes them from that area before they
get a chance to recombine. As a result, the reverse current (photocurrent) occurs. Photocurrent grows in proportion to the
intensity of light. For larger values of reverse bias voltage photocurrent does not depend on the applied voltage but is
practically constant. The number of generated electron-hole pairs increases only with the increasing of the light intensity
[7]. Moreover the designer has three basic detector choices - the silicon PIN detector, the silicon avalanche photodiode (APD)
and the photomultiplier tube (PMT) [8]. However at best, a pin diode can generate only one electron-hole pair per incident
photon and therefore a pre-amplifier is needed for low light level detection. This limitation can be overcome by using
(APDs) [2]. They operate by converting clusters of detected photons, associated with information-carrying pulses of light in
a digital communication system, into cascades of electrons. These cascades have sufficiently high charge to be readily
detected by the electronics following the APD [9]. Furthermore their high sensitivities have recently led them to be widely used
in high-radiation environments [9]-[12], which include reactors and the space environments. Therefore the radiation sources
influence the APD properties can be neutrons, gamma [13].
However the main problem is caused by the neutron radiation when the new photodetector prototypes have been used [14].
Since neutron irradiation is believed to cause large clusters of
disorder (defects) in the crystal lattice which leads to increase
dark current as well as noise and thus raise the minimum detectable light power [15]-[18].The other important permanent
damage effect is a change of the effective doping concentration (Neff) that will modify the avalanche gain behavior of the
APD. Such defects could also act as trapping centers and cause a change of APD parameters such as the change collection
efficiency and quantum efficiency and a reduction in optical sensitivity [5], [15],[19]-[21].
Thus, it is important to know the properties of the APDs in the irradiation environment, and how we can optimize these
properties in order to improve APDs characteristics. Consequently, we are concerned with radiation effects on
device efficiency, responsitivity, sensitivity, dark current and excess noise factor which enables us to calculate the
humiliation that occurs in APDs performance under irradiation environment. In addition, this allows improving the detectivity,
noise equivalent power, and signal to noise ratio which
contributes in achieving maximum usage of the communication system bandwidth and to control APDs properties in radiation
environments. The arising effects of radiation induced damage are decisive in designing high-bit-rate optical communication
systems. This work is done by using VisSim environment. VisSim is a visual block diagram language for simulation of
dynamical systems and model based design of embedded systems. It uses a graphical data flow paradigm to implement
dynamic systems based on differential equation. This paper is organized as follows: Section II presents the basic assumptions
and modeling of radiation induced damage, section III describes the model results. However section IV is devoted to
conclusion.
II. BASIC ASSUMPTIONS AND MODELING OF RADIATION
INDUCED DAMAGE
In this model two Avalanche photodiodes were used,
one of effective thickness =140 μm and the other is equal to
120μm and the photodiodes are assumed to overfill the fiber so that all mode groups are received. The cross sectional
areas diameters of the two devices are 0.8 and 0.5 mm.
Operating wavelength of the optical sources used for
transmitting data is 850 nm. To illustrate the optical
responses of avalanche photodiode in radiation environment,
the effect of the neutron radiation on the properties of
avalanche photodiodes such as Detectivity and Noise
Equivalent Power has been studied under different thermal
condition. Consequently we should reveal the harmful
neutron radiation effect that has two main contributions to
the dark current of an APD. The first of these is the surface dark current, generated at or near of the perimeter of the
junction. This current occurs outside the multiplying region,
and is therefore not multiplied. The second contribution to
the dark current is bulk current, generated within the
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1823
www.ijsetr.org
junction region, and in an APD this current component
undergoes multiplication. Thus we have [8].
DBDSD III (1) (1)
Where ID, IDS, IDB, and M are a the total dark current, the
unmultiplied surface component of the dark current, the
multiplied component of the dark current, and multiplication
factor (gain) respectively. Moreover the dark current resulted from neutron effect can be also obtained as [22].
(2)
Therefore equation (1) can be simplified as follows:
(3)
However, under neutron fluence Ф the dark current is a
volume effect, so usually the increase in the dark current
IDark neutron is written as follows [14].
(4)
where v, α are the volume of the detector, and a parameter
which has been measured by different groups, and its value
is ranging between α = (10-13)10-17A/cm. In the case of
Avalanche Photodiodes, the volume which is relevant for
the generation of the bulk current is the area times the
effective thickness of the detector [14].
(5)
Therefore equation (3) can be rewritten as follows:
(6)
Moreover one of important characteristics that are
commonly used to describe photodetector performance is
the quantum efficiency (η) that can be defined as the ratio
of number of incident photon to the face of photodiode in
unit time and is denoted as (NP) to the number of
photoexcitation photoelectron (Ne). Then the quantum
efficiency η is defined as [23]:
%100p
e
N
N (7)
Where
q
IN
phe ,
hc
PN p
(8)
where P is physical radiated power of the incident light, Iph
is the photo-current which produced in photodiode when the radiated power is P. Thus, Substitute (8) into (7), we have
[23].
%100Pq
hcI ph
(9)
Where λ, h, q, and c is incidence light wavelength, Planck
constant (6.626×10-34J·S), electric charge, and light velocity respectively. Furthermore from equation (9) we
obtain the relationship between photogenerated current Iph
and optical power P in photodiode [23]
Phc
qI ph
(10)
In addition when the wavelength of incident light (light
source) and photosensitive device (material) are fixed the
value of (ηqλ/hc) is constant. Consequently we can write the
photodiode sensor’s responsivity SR as in the following
equation [23]
hc
q
P
I ph (11)
So we can suppose that [23]
hc
qSR
(12)
Substitute SR into (10), then we obtain [23]
PSI Rph (13)
Thus, the responsivity is essentially a measure of the
effectiveness of the detector for converting electromagnetic
radiation to electrical current or voltage. Furthermore
equation (9) can be written as the following [23]
q
hcSR (14)
Moreover, the quantum efficiency η, can be also calculated
from the formula [11]
q
hcS I (15)
Where SI is the primary photoelectric sensitivity of
avalanche photodiodes which can be calculated from the
equation [11]
hc
q
P
IS
phI
(16)
Consequently, from eqs. (12), (16) we can conclude that:
RI SS (17)
However, the total sensitivity is related to the primary
sensitivity by the following equation [11]
IT MSS (18)
Furthermore, from equation (17) the total sensitivity is
related to the responsivity by the following equation
RT MSS (19)
So that responsivity gives a measure of the detector’s
sensitivity to radiant energy. Moreover another important
characteristic used to describe photodetector performance is
the signal to noise ratio (S/N ratio) that can be determined
for APDs by the equation [24]:
L
T
DBDSph
ph
R
kTBMFMIIMFMIqB
IMM
N
S
422 10
30022
22
(20)
Where k, T, B, RL, and F(M) are the Boltezman,s constant,
temperature, load resistance, and excess noise factor
respectively. However the excess noise factor may be
calculated using the model developed my McIntyre which
considers the statistical nature of avalanche multiplication.
Thus the excess noise factor is given by [8]
121 EFFEFF KKF (21)
Where KEFF is the effective k-factor for an APD, whereas k-
factor is the carrier ionization ratio. Furthermore in equation
(20) the first component in the denominator describes the
shot noise (Nsh) but the second one the thermal noise (Nth) of
the load resistance. So that from equation (20) the shot noise
can be calculated from the equation [24]
MFMIMFMIqBN DBphsh222 (22)
Moreover shot noise is present in all photon detectors due to
the random arrival rate of photons from the source of radiant
energy under measurement and background radiation. This shot noise is the true ultimate limitation to detector
performance. However, the thermal noise can be calculated
from the equation [24]
DBDarkDarkDark IprotonIgammaIneutronI
neutronIII DarkDSD
neutronIDark
dA
DSD II
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1824
www.ijsetr.org
Rb VV
V
30018.0,,
Lth
R
kTBN
4 (23) (22)
It occurs in all conducting materials. It’s a consequence of
the random motion of electrons through a conductor. The
electrons are in constant motion, but they collide frequently with the atoms or molecules of the substance. Each free
flight of an electron constitutes a minute current. The sum of
all these currents taken over a long period of time must, of
course, be equal to zero. But their ac component is thermal
noise. Therefore the total noise (NT) can be calculated from
the following equation [24]
thshT NNN (24) (23)
Moreover a substantial characteristic for an avalanche
photodiode is dependence of a multiplication factor (M) on its reverse bias voltage. Therefore an empirical relationship
describes this characteristic for parametrically variable
temperature is given by [24].
(25)
Where Vb, VR, T are the breakdown voltage, and the reverse
bias voltage, and ambient temperature respectively. In
addition an approximate universal expression of the
breakdown voltage for all semiconductors studied can be
given as follows [25].
(26)
Where Eg, Neff are the band gab energy of silicon, and the effective doping concentration respectively. Furthermore,
defects generated during irradiation cause changes in the
effective doping concentration which are macroscopically
modeled as a function of the radiation particle fluence, Ф
given by [25].
(27)
Avalanche parameters of a photodiode at a given operating
voltage (at the fixed λ and T) determine about a signal to
noise ratio, a noise equivalent power (NEP) and resulting
from it detectivity D or D*. In correctly processed photodiode structures, their noise is of shot nature. The
dependencies of total noise and dark noise on gain are [11].
(28)
Substituting from equations (2), and (4) in to equation (28).
The following equation can be obtained
(29)
Where B is the system bandwidth. Moreover based on
equation (9) the noise equivalent power can be calculated as
follows [11]. The amount of optical power incident on the
surface of a photodetector that produces a signal at the output of the detector just equal to the noise generated
internally by the detector is the noise equivalent power
(NEP). This is usually the minimum detectable signal level.
On the other hand when we compare the detectors in terms
of their detecting ability, the best detector is the one with the
lowest NEP.
MS
BINEP N
0
/ (30)
However noise equivalent power (NEP) cannot be used as
the only measure of a detector's relative performance, but
rather detector detectivity at a specific wavelength and
bandwidth should be used to determine the optimum
detector type for a given application, so if we take the
reciprocal of the NEP, we can
define the detectivity. The
higher the value of the
detectivity, the better the detector. The detectivity (D) is
given by the following equation [8].
(31)
Nevertheless the NEP and hence the value of detectivity
depends on the area of the detector. This makes it difficult to
compare the intrinsic
properties of two different
types of detectors. To remove this dependence, we
use another term, called D* (detectivity normalized to the
unit of detector area 1 cm2) to rate photodetectors. D* is
given by [11].
(32)
III. SIMULATION RESULTS AND DISCUSSIONS
We illuminate the effect of those environments on the
optical signal detecting by APDs and how we can reduce the
neutron effect in those devices. The results showed that the
key to reducing thermal radiation effect is minimizing the
active device volume and increasing device initial sensitivity
or by selecting the materials that have lower ionization ratio.
Moreover reducing the frequency band will contribute in
improving detectivity and noise equivalent power.
Furthermore optimizations of the device fabrication
processes have resulted in significantly better electrical
characteristics while maintaining good optical properties. The values of the parameters are shown in Table 1.
Table 1. List of the operating parameters used in model [1, 11, 24, 25]
Symbol Operating parameter Value
Ф Radiation Fluence 1x1011
n/cm2-5x10
12 n/cm
2
T Ambient Temperature 300 K - 400 K
P Incident Power 1 nW - 10 nW
B Bandwidth 1 MHz – 2 MHz
λ Incidence light
Wavelength 800 nm – 1100 nm
S0 Detector Initial Sensitivity 0.5 A/W – 0.75 A/W
Eg Energy gap of Silicon 1.15 eV
v Detector Volume 2.5X10-5
cm3 - 7.0X10
-5cm
3
k Boltzman's Constant 1.38x10-23
J/K
RL Load Resistance 4 KΩ - 5 KΩ
43
316
23
101.1
cm
N
eV
EaV
effgb
1515 10139.1exp1053.1effN
21
22 BIFIIqI DSDBphN
21
2 BIFqI DSN
NEP
BI
SD
N
1
2
1
0
NEP
A
BI
ASD d
N
d
2
1
0
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1825
www.ijsetr.org
Reverse Bias Voltage, VR, V
Fig.1. Variations of the multiplication factor (M) against reverse bias voltage (VR) with T=300K, P=10nW, B=1MHz, Ф=1x1011 n/cm2, λ=850nm, and RL=4.5 KΩ.
In figure (1) the typical computed values for the Multiplication Factor (M) against reverse bias voltage (VR) under thermal
effect, and experimental values are represented in the above figure that shows a good agreement between the resulted curve
and published experimental curve.
Radiation Fluence, Ф x1011, n/cm2
Fig.2. Variations of the quantum efficiency (η) against radiation fluence (Ф) at different optical signal wavelength (λ) with T=320K, P=10nW, B=1MH, and RL=4 KΩ.
Figure (2) reveals the relation between quantum efficiency
and radiation fluence at different incidence light
wavelength. The figure illustrates that as the radiation
fluence increases the device quantum efficiency is
decreases. This can be explained by the fact that if the
energy of neutron irradiation is high enough (above several hundred keV) the pair generation occurs. The energy
spectrum of the energetic electrons is called”slowing
spectrum”. The energetic electron interacts with a lattice
atom. As a result, the lattice atom is displaced from lattice
site. This is so called Primary Knock on Atom (PKA).
Interstitial (PKA), vacancy, and complex of them form a
deep level in bandgap. This is so-called the generation-
recombination centre [7]. Moreover these defects created
throughout the optically active volume tend to decrease the
overall collection efficiency of optically generated minority
carriers by reducing the minority carrier diffusion length [5].
Consequently the photocurrent decrease and hence, the
device quantum efficiency decreases. In addition quantum
efficiency decreases with increase incidence light
wavelength. This can be explained by examining some basic
material properties and the photocurrent generation mechanism. For indirect bandgap semiconductors, such as
silicon, the optical absorption coefficient is usually
relatively small for photons with energies only slightly
larger than the material bandgap. This results in the creation
of infrared (0.8 pm - 1.1 pm for Si) generated minority
carriers farthest from the front optical surface and usually
farthest from the electrical junction [5].
1.6
2.6
3.6
4.6
5.6
6.6
7.6
8.6
9.6
50 52 54 56 58 60 62 64 66 68 70
Puplished Results
Model Results
Mult
ipli
cati
on F
acto
r, M
Quan
tum
Eff
icie
ncy
, η
, %
22
27
32
37
42
47
52
57
62
67
0 1 2 3 4 5 6 7 8 9 10
η at λ= 850 nm
η at λ=1100 nm
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1826
www.ijsetr.org
Temperature, T, K Fig.3. Variations of the quantum efficiency (η) against temperature (T) at different optical signal wavelength (λ) with Ф = 1x1011 (n/cm2), P=10nW, B=1MH, and RL=4 KΩ.
Figure (3) reveals the relation between quantum efficiency
and temperature at different incidence light wavelength. The
figure shows that as the temperature increases the device
quantum efficiency decreases. This can be attributed to the
fact that temperature plays a significant role in the number
of reactions that take place during/after irradiation as a rise
in 100C can double the reaction rates [26]. Therefore this
duplication of defects created throughout the optically active
volume tends to decrease the overall collection efficiency of
optically generated minority carriers by reducing the
minority carrier diffusion length [5]. This is the reason of
why dark current increases by a double of its value for each
rise in temperature of 100C [24]. In addition, the avalanche
process statistics generate current fluctuations these
fluctuations increases with the temperature, and APD
efficiency degraded [8].
Radiation Fluence, Ф x1011, n/cm2 Fig. 4. Variations of the total sensitivity (St) against radiation fluence (Ф) at different optical signal wavelength (λ) with T=320K, P=10nW, B=1MHz, and RL=4 KΩ.
Figure (4) reveals the relation between total sensitivity and
radiation fluence. It shows the reduction in optical sensitivity with increasing neutron radiation. This is because
radiation induced defects created throughout the optically
active volume tending to decrease the overall collection of
optically generated minority carriers by reducing the
minority carrier diffusion length. The crystal damage
clusters serve as recombination sites causing optically
generated hole-electron pairs to recombine before they can
contribute to junction current [5]. In addition sensitivity is
limited by photon shot noise which increases with increase dark current in the radiation field [8]. Moreover the shorter
the wavelength of incident radiation is the higher noise
current IN and lower the gain M for IN = const. Worsening of
photodiode properties for shorter wavelengths results from
the increase of number of holes participating in total current
flowing through the avalanche region. Therefore from
Q
uan
tum
Eff
icie
ncy
, η
, (%
)
20
22
24
26
28
30
32
34
0 50 100 150 200 250 300 350 400
η at λ= 850 nm
η at λ=1100 nm
Tota
l S
ensi
tivit
y,
St ,
A/W
48
53
58
63
68
73
78
0 1 2 3 4 5 6 7 8 9 10
η at λ= 850 nm
η at λ=1100 nm
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1827
www.ijsetr.org
equation (20) this will reduce total sensitivity at shorter
wavelengths [7].
Radiation Fluence, Ф x1011, n/cm2 Fig.5. Variations of the signal to noise ratio (S/N) against radiation fluence (Ф) at different active volume (v) where (v1= 2.4x10-5 cm3, v2 = 7x10-5 cm3) with T=320K, P=10nW, B=1MHz, and RL=4 KΩ.
Figure (5) reveals the relation between signal to noise ratio
and radiation fluence. The instantaneous photocurrent
generated in photodiodes during exposure to a pulse of ionizing radiation is a result of a combination of Compton
scattering effects and the photoelectric effect, depending on
the energy of the ionizing-radiation photon. Both
mechanisms create excited electrons which generate many
electron-hole pairs as the excited electrons move through the
crystal lattice. These electron-hole pairs can contribute to
the device current if they are created within approximately
one diffusion length of the semiconductor depletion region
edge. This ionizing-radiation induced photocurrent can be
significantly larger than the intended signal current
component, depending on the amplitude of the ionizing-
radiation field and the amplitude of the signal intensity.
Moreover an increase in the photodiode dark (leakage) current results in a larger background noise level.
Consequently the signal to noise ratio decrease. However by
minimizing the active device volume, the magnitude of the
radiation induced signal current can be reduced. Relatively
short minority carrier diffusion lengths also decrease the
number of ionizing-radiation induced minority carriers that
contribute to the device current [5]. Small active volume
photodiode structures are used to achieve low noise current
generation during exposure to ionizing-radiation and as a
result increase the signal to noise ratio [6].
Temperature, T, K Fig.6. Variations of the signal to noise ratio (S/N) against temperature (T) at different load resistance (RL) with Ф = 1x1011 (n/cm2), P=10nW, B=1MHz, and RL=4 KΩ.
Figure (6) reveals the relation between signal to noise ratio
and temperature at different load resistance. From the figure
signal to noise ratio decreases as temperature increase. This
can be attributed to the fact that temperature plays a
Sig
nal
to N
ois
e R
atio
, S
/N
640
650
660
670
680
690
700
710
0 1 2 3 4 5 6 7 8 9 10
S/N at v1
S/N at v2
Sig
nal
to N
ois
e R
atio
, S
/N
50
150
250
350
450
550
0 50 100 150 200 250 300 350 400
S/N at RL= 5 KΩ
S/N at RL= 4 KΩ
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1828
www.ijsetr.org
significant role in the number of reactions that take place
during/after irradiation as a rise in 100C can double the
reaction rates [26]. Therefore this duplication of defects
created throughout the optically active volume tends to
increase photodiode dark current and therefore photodiode
noise power level [5]. So the signal to noise ratio decrease
with temperature. Moreover for the higher values of a load
resistance the value of thermal noise decreases. Hence, the
value of signal to noise ratio increase. Therefore in order to
have maximal value of a signal to noise ratio the high values
of load resistance should be taken [24].
Multiplication Factor, M
Fig.7. Variations of the signal to noise ratio (S/N) against multiplication factor (M) at different load resistance (RL) with Ф = 1x1011 (n/cm2), P=10nW, T=320 K, B=1MHz, and RL=4 KΩ.
Figure (7) reveals the relation between signal to noise ratio and multiplication factor at different load resistance. From
the figure signal to noise ratio increases as multiplication
factor increase. This can be attributed to the fact that a
mechanism of internal gain causes significant increases in a
signal current generated in a detector and improves the
signal to noise ratio. Such a mechanism, however dose not influence on the noises originated from a load resistance as
will as the amplifier noises [24]. Furthermore as thermal
noises decreases with increase the load resistance this will
contribute in increasing the signal to noise ratio of avalanche
photodiode.
[
Avalanche Photodiode Volume, v x10-5, cm3
Fig.8. Variations of the dark current (ID) against avalanche photodiode volume (v) at different radiation fluencies where (Ф1=2x1011 n/cm2, Ф2=1x1011 n/cm2) (Ф) with T=320K, P=10nW, B=1MHz, and λ=850nm.
Figure (8) reveals the relation between the dark current (ID)
and the device volume (v); from the figure it is apparent that
the dark current increase as the device volume increase.
Moreover the induced dark current is increases as the
neutron fluence increase. This result can be attributed to the
creation of the defects in the silicon lattice which are
responsible for the increase of the dark current is a volume
effect, so usually the using of APD that has large volume
will contribute in increasing of dark current ID increment
than APD that has small device volume. Moreover by
minimizing the active device volume, the magnitude of the
radiation induced signal current can be reduced. Relatively
short minority carrier diffusion lengths also decrease the
number of ionizing-radiation induced minority carriers that
Sig
nal
to N
ois
e R
atio
, S
/N
0
50
100
150
200
250
300
350
400
450
0 50 100 150 200 250 300 350 400 450 500 550 600
S/N at RL= 5 KΩ
S/N at RL= 4 KΩ
Dar
k C
urr
ent,
ID, nA
100
102
104
106
108
110
112
114
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7
ID at Ф1
ID at Ф2
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1829
www.ijsetr.org
contribute to the device current [5]. Furthermore Care must
be taken not to make the active region thickness too narrow
or the minority carrier diffusion length too small or the
desired optically generated minority carriers will not
contribute to the device current and the optical responsivity
will be suboptimal [5].
Ionization Ratio, k
Fig.9. Variations of the excess noise factor (F) against ionization ratio (k) at different temperatures (T) with P=10nW, B=1MHz, Ф=1x1011 n/cm2, λ=850nm, and RL=4 KΩ.
Figure (9) reveals the relation between the excess noise
factor F (M) and the ionization ratio (k); from the figure it is
apparent that the excess noise factor increase with
increasing the temperature and ionization ratio. This result
can be attributed to; all avalanche photodiodes generate excess noise due to the statistical nature of the avalanche
process [8]. In addition these avalanche process statistics
generate current fluctuations these fluctuations increases
with the temperature, and APD performance is degraded by
an excess noise factor (F) compared to a PIN especially at
higher temperatures [8]. Furthermore the excess noise factor
is a function of the carrier ionization ratio, k, where k is
usually defined as the ratio of hole to electron ionization probabilities [8]. Therefore the lower the values of k the
lower the excess noise factor [8].
Multiplication Factor, M Fig.10. Variations of the excess noise factor (F) against multiplication factor (M) at different effective k-factor with T=320K, P=10nW,
B=1MHz, Ф=1x1011 n/cm2, λ=850nm, and RL=4 KΩ.
Figure (10) reveals the relation between the excess noise
factor F (M) and the multiplication factor (M); from the
figure it is apparent that the excess noise factor increase
with increasing multiplication factor. This result can be accredited to the fact that all avalanche photodiodes
generate excess noise due to the statistical nature of the
avalanche process [8]. Therefore, higher multiplication
factor results in higher excess noise factor. Moreover the
excess noise factor is a function of the carrier ionization ratio, k, where k is usually defined as the ratio of hole to
Exce
ss N
ois
e F
acto
r, F
(M
)
2
2.5
3
3.5
4
4.5
5
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
F at T = 350 K
F at T = 320 K
Exce
ss N
ois
e F
acto
r, F
(M
)
0
1
2
3
4
5
6
7
8
0 10 20 30 40 50 60 70 80 90 100
F at KEFF = 0.02
F at KEFF = 0.04
F at KEFF = 0.06
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1830
www.ijsetr.org
electron ionization probabilities [8]. Therefore the lower the
values of k and M; the lower the excess noise factor [8].
Ionization Ratio, k
Fig.11. Variations of the excess noise factor (F) against ionization ratio (k) at different reverse bias voltage (VR) with T=320K, P=10nW, B=1MHz, Ф=1x1011 n/cm2, λ=850nm, and RL=4 KΩ.
Figure (11) reveals the relation between the excess noise
factor F (M) and the ionization ratio (k) at different reverse
bias voltage; from the figure the excess noise factor increase
with increasing reverse voltage. This result can be credited
to higher reverse bias voltage results in higher multiplication
factor [24]. However the avalanche gain causes an increase
in excess noise. Therefore the excess noise factor increases
with increase reverses bias voltage.
Multiplication Factor, M Fig.12. Variations of the noise equivalent power (NEP) against multiplication factor (M) at different Noise current (IN) where (IN1= 0.4 pA/Hz1/2, IN2= 0.2 pA/Hz1/2) with T=320K, P=10nW, B=1MHz, Ф=1x1011 n/cm2, λ=850nm, and RL=4 KΩ.
Figure (12) reveals the relation between noise equivalent
power and the multiplication factor. In addition based on the
definition of NEP that is the amount of optical power
incident on the surface of a photodetector that produces a
signal at the output of the detector just equal to the noise
generated internally by the detector. This is usually the
minimum detectable signal level. On the other hand a
mechanism of internal gain causes significant increases in a
signal current generated in a detector and improves the
signal to noise ratio [24]. So that NEP reduces with
increasing the multiplication factor. Furthermore in figure
(7) since noise currents which proportional to the square
E
xce
ss N
ois
e F
acto
r, F
(M
)
1.9
2
2.1
2.2
2.3
2.4
2.5
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
F at VR = 55 V
F at VR = 45 V
Nois
e E
quiv
alen
t P
ow
er, N
EP
, pW
/Hz
1/2
0.25
0.75
1.25
1.75
2.25
2.75
3.25
3.75
4.25
4.75
0 10 20 30 40 50 60 70 80 90 100
NEP at IN1
NEP at IN2
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1831
www.ijsetr.org
root of leakage currents determine the minimum detection
levels [24]. Therefore NEP value decreases with lower noise
current.
Initial Sensitivity, S0 , A/W Fig.13. Variations of the noise equivalent power (NEP) against initial sensitivity (S0) at different radiation fluencies (Ф) where (Ф1= 5x1012 n/cm2, Ф2= 1x1011 n/cm2) with P=10nW, B=1MHz, T=320K, λ=850nm, and RL=4 KΩ.
Figure (13) reveals the relation between the noise equivalent
power (NEP) and the Initial sensitivity (S0) at different
radiation fluencies, from the figure, noise equivalent power
decrease with increasing device initial sensitivity, since NEP
represents the minimum detectable signal level. However
neutron irradiation is believed to cause large clusters of
disorder in the crystal lattice due to the relatively large
energy transfer associated with neutron scattering. These
disorder clusters in the crystal lattice create generation sites
in the depletion region of the semiconductor junction. This
causes an increase in the depletion layer generation
recombination current. This results in an increase in the
photodiode dark (leakage) current and therefore a larger
background noise level [5]. Furthermore this tends to raise
the minimum power level of detectable optical signals. So
this leads to NEP increment [6].
Reverse Bias Voltage, VR, V Fig.14. Variations of the detectivity (D) against reverse bias voltage (VR) at different initial sensitivity (S0) with P=10nW, B=1MHz, Ф=1x1011 n/cm2, λ=850nm, and RL=4 KΩ.
Figure (14) reveals the relation between detectivity (D) and
the Reverse bias voltage (VR) at different initial sensitivity
(S0); from the figure it is apparent that the detectivity increase with increasing reverse bias voltage. For the reason
that reverse bias voltage results in increasing multiplication
factor which causes significant increases in a signal current
generated in a detector and improves the signal to noise ratio [24]. Furthermore this result in reducing the minimum
N
ois
e E
quiv
alen
t P
ow
er, N
EP
, pW
/Hz1
/2
1.6
2.6
3.6
4.6
5.6
6.6
7.6
8.6
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
NEP at Ф1
NEP at Ф2
Det
ecti
vit
y,
D x
10
5, cm
Hz1
/2W
-1
0.06
0.11
0.16
0.21
0.26
0.31
0.36
0 5 10 15 20 25 30 35 40 45
D at S0 = 0.75 A/W
D at S0 = 0.5 A/W
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1832
www.ijsetr.org
power level of detectable optical signals [20]. Consequently
detectivity will increase.
Bandwidth, B, MHz Fig.15. Variations of the detectivity (D) against bandwidth (B) at different radiation fluencies (Ф) where (Ф1= 5x1012 n/cm2, Ф2= 1x1011 n/cm2) with P=10nW, T=320K, Ф=1x1011 n/cm2, λ=850nm, and RL=4 KΩ.
Figure (15) reveals the relation between detectivity (D) and
the bandwidth (B) at different radiation fluencies (Ф); from
the figure it is apparent that the detectivity decrease with
increasing bandwidth. This result can be endorsed to while
impact ionization can increase the APDs detectivity it still
limits their bandwidth at high M because of the long
multiplication chains, which takes time to build up and die
away. This accompanying increase in response time with M
results in a roughly constant gain bandwidth product [2].
While increasing the avalanche gain the detectivity increase
it causes both, an increase in the noise originated from the
dark current and in the quantum noise this restricts the
bandwidth [24].
Temperature, T, K
Fig.16. Variations of the detectivity (D) against temperature (T) at different initial sensitivity (S0) with P=10nW, B=1MHz, Ф=1x1011
n/cm2, λ=850nm, and RL=4 KΩ.
Det
ecti
vit
y, D
x 1
05, cm
Hz1
/2W
-1
1.3
1.32
1.34
1.36
1.38
1.4
1.42
1.44
1.46
1.48
1.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
D at Ф1
D at Ф2
Det
ecti
vit
y, D
x 1
05, cm
Hz1
/2W
-1
0
1
2
3
4
5
6
190 240 290 340 390
D at S0 = 0.75 A/W
D at S0 = 0.5 A/W
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1833
www.ijsetr.org
Figure (16) reveals the relation between detectivity (D) and
the temperature (T) at different initial sensitivity (S0); from
the figure it is apparent that the detectivity decrease with
increasing temperature. For the reason that rises in 100C can
double the reaction rates that take place during/after
irradiation [26]. Therefore this duplication of defects created
throughout the optically active volume tends to decrease the
overall collection of optically generated minority carriers by
reducing the minority carrier diffusion length and increase
the dark current [5]. Moreover the avalanche process
statistics generate current fluctuations these fluctuations
increases with the temperature [24]. As a result this will
raise the minimum optical power an optical detector can
sense and reduce device detectivity [5].
Noise Current, IN , pA/Hz1/2 Fig.17. Variations of the normalized detectivity (D*) against noise current (IN) at different Initial sensitivity (S0) with T=320K, P=10nW,
B=1MHz, Ф=1x1011 n/cm2, λ=850nm, and RL=4 KΩ.
Figure (17) reveals the relation between normalized
detectivity (D*) and the Noise current (IN); from the figure
it is apparent that the normalized detectivity decrease with
both increasing noise current and decrease the initial
sensitivity. This result can be attributed to increasing the noise current will reduce signal to noise ratio and tends to
impose a lower bound on the minimum detectable optical
signal for a given signal to noise ratio [6]. Therefore
higher noise current results in lower normalized
detectivity. Moreover decrease the initial sensitivity result
in raising the minimum power level of detectable optical
signals [20], hence the normalized detectivity decrease.
IV. CONCLUSION
In this paper a block diagram model treating the
radiation induced damage is proposed to provide a mean to
control the optical properties of APDs in thermal radiation environments. The proposed treatment can be used to
improve the detectivity and noise equivalent power as well
as the normalized detectivity. A model is built using VisSim
environment. The obtained results showed that the key to
reducing thermal radiation effect would expect in
minimizing the active device volume. In the same time this
provides higher detector sensitivity and responsitivity as
well as detector efficiency. Moreover increase signal to
noise ratio can be achieved by increase the load resistance.
Furthermore increasing reverse bias voltage results in
increase avalanche gain multiplication factor. Moreover reducing the frequency band will contribute in improving
detectivity and noise equivalent power. Furthermore this can
be achieved by selecting the materials that have lower
ionization ratio through the fabrication process. These
results are significant for many applications in which noise
currents which proportional to the square root of leakage
currents determine the minimum detection levels. The
results are validated against published experimental work in
temperature case and show good agreement.
V. REFERENCES
[1] M. Kovacevic, S. Savovic, A. Djordjevich, J. Bajic, D. Stupar , M. Kovacevic, and S. Simic, "Measurements of
growth and decay of radiation induced attenuation during the irradiation and recovery of plastic optical fibres",
Optics and Laser Technology, Vol. 47, pp. 148–151, 2013. [2] C. Tan, J. Ng, S. Xie, and J. David, "Potential materials
for avalanche photodiodes operating above 10Gb/s", International Conference on Computers and Devices for
Communication, Vol. 9, pp. 152-157, 2009. [3] A. Kalma, and W. Hardwick, "Radiation testing of pin
photodiodes ", IEEE Transaction on Nuclear Science, Vol. 25, pp. 1483-1488, 1978.
N
orm
aliz
ed D
etec
tivit
y, D
* x
10
5, cm
Hz1
/2W
-
1 0
1
2
3
4
5
6
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
D* at S0 = 0.75 A/W
D* at S0 = 0.5 A/W
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1834
www.ijsetr.org
[4] S. M. Eladl, "Modeling of ionizing radiation effect on
optoelectronic-integrated devices (OEIDs)", Microelectronics Journal, Vol. 40, pp.193–196, 2009.
[5] J. Wiczer, T. Fischer, L. Dawson, G. Osbourn, T. Zipperian, and C. Barnes, "Pulsed irradiation of optimized,
mbe grown, AlGaAs/GaAs radiation hardened photodiodes", IEEE Transaction on Nuclear Science, Vol.
31, pp. 1477-1482, 1984. [6] J. Wiczer, L. Dawson, G. Osbourn, and C.
Barnes,"Permanent damage effects in si and AlGaAs/GaAs photodiodes", IEEE Transaction on Nuclear Science, Vol.
29, pp. 1539-1544, 1982. [7] D. Nikolic, A. Vasic, I. Fetahovic, K. Stankovic, and P.
Osmokrovic, "Photodiode Behavior in Radiation Environment", Application Mathematics Information and
Mechanics, Vol. 3, pp. 27-34, 2011. [8] Avalanche Photodiodes: A User's Guide, PerkinElmer,
2003.
[9] M. C. Teich, K. Matsuo, and B. E. A. Saleh, "Counting distributions and error probabilities for optical receivers
incorporating superlattice avalanche photodiodes", IEEE Transactions on Electron Devices, Vol. 33, pp. 1475-1488,
1986. [10] J. S. Laird, S. Onoda, T. Hirao, L. Edmonds, and T.
Ohshima, "Quenching of impact ionization in heavy ion induced tracks in wide bandwidth si avalanche
photodiodes", IEEE Transaction on Nuclear Science, Vol. 9, pp. 1704-1709, 2007.
[11] I. Wegrzecka, M. Wegrzecki, M. Grynglas, J. Bar, A. Uszynski, R. Grodecki, P. Grabiec, S. Krzeminski, and T.
Budzynski, "Design and properties of silicon avalanche photodiodes", Opto-Electronics Review, Vol. 12, pp. 95–
104, 2004. [12] J. S. Laird, T. Hirao, S. Onoda, H. Ohyama, and T.
Kamiya, "Heavy-ion induced single-event transients in high-speed InP-InGaAs avalanche photodiodes", IEEE
Transaction on Nuclear Science, Vol. 50, pp. 2225-2232, 2003.
[13] R. H. West, "A local view of radiation effects in fiber optics", Journal of Lightwave Technology, Vol. 6, pp. 155-
164, 1988. [14] S. Baccaro, J. Bateman, F. Cavallari, V. Ponte, K. Deiters,
P. Denes, M. Diemoz, T. Kirn, A. Lintern, E. Longo, M. Montecchi, Y. Moussienko, J. Pansart, D. Renker, S.
Reucroft, G. Rosi, R. Rusack, D. Ruuska, R. Stephenson, and M.J. Torbet, "Radiation damage effect on avalanche
photodiodes", CMS Conference, Vol. 10, pp. 1-7, 1998.
[15] H. Lischka, H. Henschel, W. Lennartz, and H. U. Schmidt, "Radiation sensitivity of light emitting diodes
(LED), laser diodes (LD) and photodiodes (PD)", IEEE Transaction on Nuclear Science, Vol. 91, pp. 404-408,
1992. [16] A. Niemela, H. Sipila, and V.I. Ivanov, "High-resolution
p-i-n CdTe and CdZnTe X-Ray detectors with cooling and rise-time discrimination", IEEE Transactions on Electron
Devices, Vol. 30, pp. 549-553, 1996. [17] S. Onoda, T. Hirao, J. S. Laird, T. Wakasa, T. Yamakawa,
T. Okamoto, Y. Koizumi, and T. Kamiya, "Development of monte-carlo modeling for proton induced charge in si
pin photodiode", IEEE Transaction on Nuclear Science, Vol. 51, pp. 2770-2775, 2004.
[18] L. Shi, S. N. Nihtianov, F. Scholze, and L. K. Nanver, "Electrical performance stability characterization of high-
sensitivity si-based EUV photodiodes in a harsh industrial application", IEEE Transaction on Nuclear Science, Vol. 2,
pp. 3952-3957, 2012.
[19] T. English, G. Malley, and R. Korde, "Neutron hardness
of photodiodes for use in passive rubidium frequency standards", IEEE Annual Frequency Control Symposium,
Vol. 2, pp. 532-539, 1988. [20] K. Gill, V. A. Engels, J. Batten, G. Cervelli, R. Grabit, C.
Mommaert, G. Stefanini, J. Troska, and F. Vasey, "Radiation Damage Studies of Optoelectronic Components
for the CMS Tracker Optical Links", IEEE Transaction on Nuclear Science, Vol. 98, pp. 405-412, 1998.
[21] J. Jiménez, J. S. Páramo, M. Álvarez, J. Domínguez, J. Oter, I. Arruego, R. Tamayo, and H. Guerrero, "Proton
radiation effects on medium/large area Si PIN photodiodes for Optical Wireless Links for Intra-Satellite
Communications (OWLS)", IEEE Transaction on Nuclear Science, Vol. 7, pp. 73-79, 2007.
[22] Ahmed Nabih Zaki Rashed, "Performance Response Characteristics of Avalanche Photodiodes (APDs) Under
High Thermal and Protons Irradiation Environments",
Global Advanced Research Journal of Engineering, Technology and Innovation, Vol. 1, pp. 033-042, 2012.
[23] Z. SU, and X. Liang, "Computation and analysis on The volt-ampere characteristics of photodiode sensor under the
certain conditions", IEEE International Congress on Image and Signal Processing, Vol. 7, pp. 2593-2596,2011.
[24] Z. Bielecki, "Analysis of operation conditions of avalanche photodiodes on signal to noise ratio", Opto-
Electronics Review, Vol. 5, pp. 249–256, 1997. [25] A. A. Mohamed, M. El-Halawany, Ahemed Nabih Zaki
Rashed, and H. El-Hageen, "Harmful proton radiation damage and induced bit error effects on the performance of
avalanche photodiode devices", International Journal of Multidisciplinary Sciences and Engineering, Vol. 2, pp.
27-36, 2011. [26] K. A. Murray, J. E. Kennedy, B. M. Evoy, O. Vrain, D.
Ryan, D. Ryan, and C. L. Higginbotham, "Effects of gamma ray and electron beam irradiation on the
mechanical, thermal, structural and physicochemical properties of poly (ether-block-amide) thermoplastic
elastomers", Journal of the Mechanical Behavior of Biomedical Materials, Vol. 6, pp. 255-305, 2012.
ISSN: 2278 – 7798 International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 2, Issue 10, October 2013
1835
www.ijsetr.org
Author’s Profile Dr. Ahmed Nabih Zaki Rashed was born in
Menouf city, Menoufia State, Egypt country in
23 July, 1976. Received the B.Sc., M.Sc., and Ph.D. scientific degrees in the Electronics and Electrical Communications Engineering Department from Faculty of Electronic Engineering, Menoufia University in 1999, 2005, and 2010 respectively. Currently, his job carrier is a scientific lecturer in Electronics and Electrical Communications Engineering
Department, Faculty of Electronic Engineering, Menoufia university, Menouf. Postal Menouf city code: 32951, EGYPT.
His scientific master science thesis has focused on polymer fibers in optical access communication systems. Moreover his scientific Ph. D. thesis has focused on recent applications in linear or nonlinear passive or active in optical networks. His interesting research mainly focuses on transmission capacity, a data rate
product and long transmission distances of passive and active optical communication networks, wireless communication, radio over fiber communication systems, and optical network security and management. He has published many high scientific research papers in high quality and technical international journals in the field of advanced communication systems, optoelectronic devices, and passive optical access communication networks. His areas of interest and experience in optical communication systems, advanced optical communication networks, wireless optical access
networks, analog communication systems, optical filters and Sensors, digital communication systems, optoelectronics devices, and advanced material science, network management systems, multimedia data base, network security, encryption and optical access computing systems. As well as he is editorial board member in high academic scientific International research Journals. Moreover he is a reviewer member and editorial board member in high impact scientific research international journals in the field of
electronics, electrical communication systems, optoelectronics, information technology and advanced optical communication systems and networks. His personal electronic mail ID (E-mail:[email protected]). His published paper under the title "High reliability optical interconnections for short range
applications in high speed optical communication systems" has achieved most popular download articles in Optics and Laser Technology Journal, Elsevier Publisher in year 2013.